There continues to be uncertainty regarding the optimal manner in which ECT should be delivered, in terms of both therapeutic and adverse effects. We have already observed that the use of unilateral nondominant (ULND) electrode placement and brief pulse (P) stimuli are each associated with lower amounts of memory dysfuntion and EEG abnormalities than traditional bilateral (BL) electrode placement and sine wave (S) stimuli. On the other hand, UNLD and P ECT modalities appeared to show remission rates equivalent to that of their counterparts. Our previous work has encouraged us to extend our focus towards the investigation of a third stimulus-related factor, stimulus intensity, which, may also have a significant influence upon both therapeutic and adverse effects. The proposed study will involve two components. The first, and most major, component will be prospective comparison of two stimulus intensity dosing strategies. Major test measures will include remission rates, onset of clinical response, memory function, and EEG slowing. The design will incorporate a three month follow-up period. Subjects will be psychiatric inpatients who have been clinically referred for ECT treatment of a major depressive episode. A stimulus intensity titration procedure will be carried out at the first ECT treatment in order to arrive at an estimate of seizure threshold, since this measure can vary widely across individuals. Because of our belief that both therapeutic and adverse effects may be related to the extent of ictal generalization during ECT, a variety of measurements will be made at the time of the ECT treatments themselves, including multilead ictal EEG monitoring and serum prolactin levels. As have been our practice, EEG data will be subjected to both visual ratings and computer analysis. Since EEG seizure patterns appear to vary more widely between individuals than they do within the same person, a second smaller component of this protocol will investigate stimulus intensity effects on an intraindividual basis. Test parameters will be as noted above for measurements made at the time of ECT treatments.